Marketing Skills for AI Agents
A collection of AI agent skills for CRO, copywriting, SEO, analytics, and growth engineering.
- Source repo
- coreyhaines31/marketingskills
- Stars
- ★ 55k
- Last updated
- 3d ago
- License
- MIT
- Primary language
- JavaScript
- FA score
- 40/100 · Major gaps
At a glance
- How it runs
- Works with
- Universal · cross-platformCodex · Claude Code
- Cost
- Free, no paid service needed
- Setup effort
- Low · running in minutes
- You'll need
- Typical use
- A SaaS founder wants to optimize conversion rate on a landing page, asking the agent to use the cro skill.
- Not a fit if
- Teams using agents that don't support the Agent Skills spec
- Users who want zero setup and no product context file to maintain
- Teams needing executable code rather than Markdown instructions
- Source review
- 40/100 · Major gaps 8 safety controls not found
What does this agent do, and when should you use it?
This repository provides a set of Markdown-based 'skills' for AI coding agents to perform marketing tasks such as conversion optimization, copywriting, SEO, analytics, and growth engineering. It follows the Agent Skills specification, making it compatible with Claude Code, OpenAI Codex, Cursor, Windsurf, and any agent supporting the spec. The core is the 'product-marketing' skill, which holds product context and is read by all other skills first. There are over 50 skills, including cro, copywriting, seo-audit, analytics, emails, and ads. Installation is flexible via the npx skills CLI, Claude Code plugin, or manual copying. The repository is maintained by Corey Haines, a conversion optimization specialist.
It provides a set of skills that AI agents can invoke. Each skill is a Markdown file with specialized knowledge and workflows for a specific task. Agents call these skills when handling marketing requests, such as optimizing landing pages (cro), writing copy (copywriting), performing SEO audits (seo-audit), setting up analytics (analytics), creating email sequences (emails), and more. Skills reference each other and build on shared context from 'product-marketing'. After installation, users can trigger skills via natural language requests or direct commands like /cro. The repository itself doesn't execute code but provides instructions for agents.
- A SaaS founder wants to optimize conversion rate on a landing page, asking the agent to use the cro skill.
- A marketer needs homepage copy for their product, using the copywriting skill.
- A developer wants to set up GA4 event tracking, using the analytics skill.
- A B2B sales rep wants to create a cold email sequence, using the cold-email skill.
- A consultant needs a comprehensive marketing plan for a client, using the marketing-plan skill.
How do you install or deploy this agent?
Recommended via npx skills CLI:
npx skills add coreyhaines31/marketingskillsInstall specific skills:
npx skills add coreyhaines31/marketingskills --skill cro copywritingFor Claude Code, use the plugin system:
/plugin marketplace add coreyhaines31/marketingskills
/plugin install marketing-skillsAlternatively, clone and copy:
git clone https://github.com/coreyhaines31/marketingskills.git
cp -r marketingskills/skills/* .agents/skills/How do you use this agent?
After installation, ask your agent for marketing tasks. For example:
- "Help me optimize this landing page for conversions" → uses cro skill
- "Write homepage copy for my SaaS" → uses copywriting skill
- "Set up GA4 tracking for signups" → uses analytics skill
- "Create a 5-email welcome sequence" → uses emails skillYou can also invoke skills directly with commands like /cro, /emails, /seo-audit.
What are this agent's strengths and limitations?
- Comprehensive coverage across marketing domains: CRO, SEO, copy, analytics, sales enablement.
- Based on Agent Skills spec, compatible with multiple AI agents (Claude Code, Codex, Cursor, Windsurf).
- Skills are interconnected with a dependency map, enabling collaborative workflows (e.g., customer-research feeds copywriting and CRO).
- Multiple installation methods (CLI, plugin, manual) offer flexibility.
- Relies on the Agent Skills spec, so only works with agents that support it.
- Requires manual maintenance of the product-marketing context file for best results.
- Skills are Markdown instructions, not executable code, so effectiveness depends on the agent's interpretation.
- Some skills may become outdated; users need to update periodically.
How does this agent compare with similar options?
Key facts side by side with the most closely related agents.
| Agent | Source review | Form / cost | Stars | Updated | Language | Full support on |
|---|---|---|---|---|---|---|
| Marketing Skills for AI Agents This agent | 40 · Major gaps | CLIFree | ★ 55k | 3d ago | JavaScript | Codex · Claude Code |
| SEO Skills for AI Agents | 27 · Major gaps | — | ★ 537 | 12d ago | — | Codex · Claude Code |
| SuperSEO Skills | 49 · Major gaps | Agent plugin / skillFreemium | ★ 360 | 1mo ago | — | Claude Code · Claude.ai |
| LinkedIn Skills for Claude Code and Codex | 61 · Some gaps | Agent plugin / skillFree + model costs | ★ 4.9k | 4d ago | Python | Codex · Claude Code · Claude.ai |
How does FollowAgents rate this agent?
Why each dimension lost points
Evidence shows the repository contains only skill definitions and documentation, with no executable code or permission requests. Therefore, least privilege, user confirmation, data flow transparency, sensitive data handling, dependency security, external effects, rollback, and source attribution are all not addressed, scoring 0. No malicious behavior found, but no security mechanisms or transparency either.
Self-consistency: README skill list matches directory structure and provides rename mapping, indicating good internal consistency, score 2. Dependency availability: Depends on external tools like npx skills and GitHub Actions, but no version pinning or mirrors, score 1. Failure messages: No error handling or user guidance provided, score 0.
Audience and scenarios: Clearly targets technical marketers and founders, covering many marketing scenarios, score 2. Capability boundaries: Each skill has a description but no explicit limits or boundaries, score 2. Trigger precision: Skill descriptions include trigger conditions but no precise trigger rules, score 2. Environment fit: Supports multiple agents and installation methods, score 2.
Information architecture: Clear directory structure and skill categories, score 2. Install notes: Multiple installation methods provided, score 2. Naming stability: Rename mapping provided, score 2. Examples and FAQ: Usage examples provided but no FAQ, score 2. Known limitations: Not explicitly listed, score 1. License: MIT license, score 2. Versioning and changelog: Upgrade guide provided but no formal changelog, score 2. Maintenance responsibility: Author and contribution guidelines clear, score 2.
Output usability: Skills provide structured outputs, score 2. Marginal value: Provides a large set of marketing skills with practical value, score 2. Cost benefit: Free and open source, easy installation, score 2.
Claim traceability: Claims in README lack specific evidence, score 1. Cross-source corroboration: No external validation, score 1. Fact-inference separation: Not clearly distinguished, score 1.
- Not found in source: least-privilege scopingGrant only what the task needs: a dedicated account or read-only token, scoped to specific directories and repos.
- Not found in source: confirmation before actingTurn on (or add) a confirmation step before it acts, and try it in a sandbox or test environment before real data.
- Not found in source: data-flow disclosureWatch which external services it contacts (proxy or firewall logs) and keep sensitive data out until you know where it goes.
- Not found in source: sensitive-data handlingUse dedicated, low-privilege, revocable API keys — never production credentials — and keep secrets out of logs.
- Not found in source: dependency securityPin versions and run a dependency audit (npm audit, pip-audit) before installing; prefer running it in a container.
- Not found in source: disclosed external effectsEstablish which external systems it writes to, sends to or changes, and verify with test accounts or repos before production.
- Not found in source: rollback or recovery pathBack up first, or work on a git branch or snapshot, so its changes can be undone.
- Not found in source: verifiable attributionInstall from the official repo or registry and check the publisher and URL to avoid look-alike packages.
- The repository contains only skill definitions with no security mechanisms or permission controls; users must assess risks themselves.
- Dependencies on external tools and third-party actions are not version-pinned, posing supply chain risks.
- No failure handling or error messages are provided; users may lack guidance when issues arise.
Recent releases
Latest entries from the project's GitHub Releases. Notes are quoted from upstream as written, not rewritten or tested by FollowAgents.
-
v2.11.37 v2.11.37 — Copywriting and ads now read the learnings ledger added to product-marketing in 2.11.34.
Release notes excerpt
Copywriting and ads now read the learnings ledger added to product-marketing in 2.11.34. - copywriting (2.3.0 → 2.3.1): checks the context file's Learnings section before drafting, so tested angles and claims aren't proposed again as new. - ads (2.6.0 → 2.6.1): reads the context file's Paid Ads section (targets, floors, conversion definitions, approval boundaries) and its Learnings section before proposing tests. Skill ZIP assets Complete source catalog: 50 skills. Claude web target: 50 skills with reviewed short descriptions. …View on GitHub ↗ -
v2.11.36 v2.11.36 — Newsletter sponsorships and partner link tracking, adapted from Wispr Flow's MIT skills…
Release notes excerpt
Newsletter sponsorships and partner link tracking, adapted from Wispr Flow's MIT skills in the skills.marketing library. - influencer-marketing (1.3.0 → 1.4.0): references/newsletter-sponsorships.md covers vetting a newsletter, reading earlier placements before booking again, skipping angles run on the same list in the last 30 days, placement specs, one file and one tracked link per placement, and reporting. Partner links now point to the UTM patterns in analytics. …View on GitHub ↗ -
v2.11.35 v2.11.35 — Copy skills now diagnose what they were handed before changing it, adapted with credit f…
Release notes excerpt
Copy skills now diagnose what they were handed before changing it, adapted with credit from Profound's and Wispr Flow's MIT copy skills. - copywriting (2.2.0 → 2.3.0): new Diagnose the Situation First section. Blank or slop input gets rebuilt from the brief, a draft with a real voice gets restored in that voice, and a point edit changes only what was asked. It reads taste from the input (vocabulary, emphasis, scope, angle). A drafting protocol adds decompose-and-lock, a three-question gate (can the reader picture it, can it be proven, could a competitor write it), rationed punch, and …View on GitHub ↗